If you want to create the store of the future, those dusty servers in the corner of the store need a refresh. The arrival of AI for retail and edge computing represents an incredible opportunity for brick and mortar stores to deliver next level customer service. In this blog post, I’ll explore how retailers have begun to deploy artificial intelligence solutions, paired with edge technology, to modernize and for operational efficiency. Let’s dive in!
Meeting shopper expectations in-store
Since the rise of the first department stores, mail order catalogs, and discount merchants, retail has been in a constant cycle of change and disruption. Over the last few decades, the emergence of online shopping meant tremendous pressure on traditional retailers to keep up. Yet 80% of all goods are still purchased the old fashioned way–in person, in a physical store.
The high share of in-store vs. online revenue doesn’t mean that brick and mortar retailers can be complacent about the in-store experience. Digital commerce has left shoppers craving some of the benefits of online browsing while in physical stores. Customers desire shopping trips that are richer, with more immersive experiences that include dynamic promotions and greater ease finding items. In addition, retailers are rapidly ideating on ways to improve associate safety and reduce theft.
Smart adoption of AI holds the promise of improving the customer experience, lifting revenue, and improving security. AI use cases that are actively being piloted by retailers include store layout and traffic flow optimization, shelf inventory scanning, voice ordering, dynamic point of sale promotions, spill detection, threat detection, and immersive shopping.
Existing store systems inhibit innovation
The retail industry has come to rely on the cloud for AI model training and inference. However, in the case of retail stores, there are some very specific challenges that prohibit the use of cloud for these functions. Retail locations, especially in rural areas, are known to have limited or unstable network bandwidth availability from the store to their own data centers or out to the cloud.
AI is far from the first use case to come up against the network limitations of a retail environment. Intermittent network connections have threatened systems that are central to a retailer’s ability to collect revenue–point of sale (POS) systems. Retailers ensure continuous processing of the POS and other critical systems by placing compute gear directly in stores. This approach of placing compute and storage close to the end-user and data sources is known as edge computing and it is uniquely suited to meet the high bandwidth, low latency requirements for AI use cases.
Why legacy stacks can be a roadblock to innovation
Today, almost every retailer already has a compute and storage tech stack sitting in a closet, under the manager’s desk or in the back store room. In many cases, however, the systems that retailers use in-store are built on legacy stacks that are a roadblock to innovation.
These systems lack modern application platform capabilities to support today’s rapid software development cycles. Existing systems are also difficult to manage centrally and inhibit achieving configuration and security policy consistency at scale. And they have limited means to expand to support AI applications.
Intelligent in-store systems at the edge of the retailer’s network
The desire to leverage AI to create engaging, real-time, immersive in-store experiences; protect against immediate safety threats; and offer innovative voice- and vision-based commerce is raising demands on retail edge computing systems. This doesn’t mean that we can afford to place the full power of cloud computing inside a retail store. That approach would be cost and space prohibitive.
A modern store architecture includes training AI models in the cloud, then pushing those models to a store platform for inference at the edge. In this way, the systems at the edge become an extension of the power of cloud capabilities without needing to be connected to the cloud continuously. Google Distributed Cloud offers a compelling set of capabilities for any organization looking to get the most from edge technology.
In essence, a slice of the cloud is placed inside the store. As an added benefit, these cloud connected systems can be operated and managed as a fleet, with centralized configuration and security policy management along with the ability to push updates and customize policies to the full fleet or subsets as it makes sense.
A modern architecture that radically simplifies the adoption curve
AI-ready store systems that support a modern application platform, centralized management, and cloud connectivity are emerging to enable innovation. These cost-effective systems offer a small footprint and an efficient, fleet-based management model. Support is available for networked, cloud connected operation as well as disconnected processing when necessary.
While all of this sounds like an ideal answer for the latest AI applications and use cases, how can a retailer practically move forward when a majority of the store applications rely on the old tech stack? Running two tech stacks in a store is untenable due to cost, physical space constraints, and additional administrative burden.
Modern edge architectures reduce operational costs
SADA has been leading the way to address this challenge, offering services that help retailers pilot and test modern edge architectures and platforms that can accommodate their transitional requirements. Designed with an ability to execute legacy VM-based applications side by side with newer containerized workloads and AI-driven applications, these edge solutions dramatically reduce the complexity of the adoption curve.
Retailers can immediately start reaping the innovation available from an AI-enabled solution without waiting to modernize all of their existing applications. Benefits also include faster development and deployment cycles for all software, more rapid pipelining of new features, and reductions to operational costs.
Edge systems can be deployed as a fully managed hardware and software service or hosted on existing or new customer-owned and -operated hardware.
The store of the future
The store of the future includes technology that drives new levels of customer engagement, fraud detection, process improvements, employee safety, inventory management automation and more. To get there, the application and platform modernization that has been applied to datacenter systems now needs to be applied to retail store systems that sit at the edge.
These hybrid cloud systems offer a lower admin burden, resilience by design, support for more frequent software releases, and single platform for all workload types. By modernizing, retailers gain the cost efficiency benefits of cloud-based, centralized management and ability to create innovative store experiences for shoppers and associates.
Start your modern edge architecture strategy with SADA
At SADA, we’re thrilled to play a part in driving AI innovation in every industry, with retail leading the adoption of GenAI and edge computing solutions. Contact us today to schedule a discovery call to start exploring how your retail business can embrace the latest in GenAI and edge innovations.